Data Indexing Methods

Chainsight supports three main ways to bring data into the network, each balancing cost, security, and speed.

A. Cost-Friendly Backend Ingestion

• Summary: A simpler approach where Chainsight’s internal nodes fetch data from an API and push it into the indexer. Minimal overhead.

• Pros: Cheaper, faster to set up. Great for frequent updates on less critical data.

• Cons: Doesn’t provide advanced cryptographic proofs of authenticity. Best used if cost is the primary concern and you trust the data source or the aggregator.

B. zkTLS-Based Ingestion

• Summary: Collaborate with an external cryptographic provider that fetches data, then proves it hasn’t been tampered with by generating a zero-knowledge proof.

• Pros: Offers high integrity. Even though the node runs off-chain, the proof ensures data authenticity.

• Cons: Range of supported data might be smaller; also typically higher overhead. Perfect for real-time pricing with strong security requirements.

C. Distributed HTTPS Outcalls (Consensus)

• Summary: A set of distributed nodes each call the same HTTPS endpoint, then reach consensus on the result. If a majority or threshold agrees, the final data is stored.

• Pros: No single aggregator. Good for widely used sources like public APIs.

• Cons: Must wait a few seconds for consensus, so slower than direct push.